The Influence of Customer Relationship Management to Customer Satisfaction and Retention in Propery and Casualty Insurance

Transcription

1 Cleveland State University ETD Archive 2008 The Influence of Customer Relationship Management to Customer Satisfaction and Retention in Propery and Casualty Insurance Brooke Ellen Lyttle Cleveland State University Follow this and additional works at: Recommended Citation Lyttle, Brooke Ellen, "The Influence of Customer Relationship Management to Customer Satisfaction and Retention in Propery and Casualty Insurance" (2008). ETD Archive. Paper 701. This Thesis is brought to you for free and open access by It has been accepted for inclusion in ETD Archive by an authorized administrator of For more information, please contact

2 THE INFLUENCE OF CUSTOMER RELATIONSHIP MANAGEMENT TO CUSTOMER SATISFACTION AND RETENTION IN PROPERTY AND CASUALTY INSURANCE BROOKE ELLEN LYTTLE Bachelor of Arts in Psychology and Criminal Justice Kent State University May, 2003 Submitted in partial fulfillment of requirements for the degree MASTER OF ARTS IN PSYCHOLOGY at the CLEVELAND STATE UNIVERSITY May, 2008

3 This thesis has been approved for the Department of Psychology and the College of Graduate Studies Thesis Chairperson, Steven D. Slane, PhD Department & Date Brian F. Blake, PhD Department & Date Leslie E. Fisher, PhD Department & Date

4 THE INFLUENCE OF CUSTOMER RELATIONSHIP MANAGEMENT TO CUSTOMER SATISFACTION AND RETENTION IN PROPERTY AND CASUALTY INSURANCE BROOKE ELLEN LYTTLE ABSTRACT Customer relationship management (CRM) emerged in the 1990 s, promising to revolutionize the business and customer dynamic. At present, CRM has yet to live up to its promise of individualized customer relationships with carefully targeted customers. In property and casualty insurance, customer and insurer relationships are important. It is more cost effective to retain current customers than to acquire new ones. This thesis explores the history of CRM and how its proper implementation can help identify areas of customer satisfaction and retention in the property and casualty insurance industry. Data were collected from a regional property and casualty insurer and analyzed to determine customer satisfaction standards. A factor analysis and several multiple regressions were conducted to determine whether satisfaction on identified standards was a predictor of stated likelihood to renew the policy. The overall regression examined independent variables under the control of the insurance company and showed a significant overall prediction, with 48.0 percent of the variance explained. When looking at the significant unique contributors, satisfaction with premium/policy factor had the greatest influence, followed closely by people service factor and claims service factor. iii

5 The second regression was conducted with customers of high-value agencies and explored variables under control of the agent. The model explained 33.8 percent of the variance, and found satisfaction with the agent had the greatest influence, followed by ease of billing, and satisfaction with explanations of premium costs. The third regression looked at the same variables but with customers of low-value agents. The model explained 47.4 percent of the variance, and found ease of the claims process had the most influence, followed by satisfaction with explanations of premium costs, and ease of billing. The goal was to investigate how variables identified through previous research would predict likelihood to renew with the insurer. The results of all the regressions support the importance of CRM moments of truth. In addition, the results from the analyses if customers of low- and high-value agents provided support for the impact of the company s internal program, FOCUS. iv

6 TABLE OF CONTENTS Page ABSTRACT iii LIST OF TABLES.. vi LIST OF FIGURES vii CHAPTER I. INTRODUCTION Database marketing.. 2 Building a database... 3 Types of databases Theoretical stages of CRM... 5 CRM process Key Moments of truth... 8 Customer Lifetime Value Hypothesis development. 14 II. METHOD Company history Survey procedure Participants III. RESULTS IV. DISCUSSION REFERENCES APPENDIX v

7 LIST OF TABLES Table Page I. Results from likelihood to renew regression of satisfaction variables. 20 II. Results from factor analysis of independent variables. 22 III. Results from likelihood to renew regression with factor scores IV. Results from likelihood to renew regression of satisfaction variables (customers of high-value agents) V. Results from likelihood to renew regression of satisfaction variables (customers of low-value agents) vi

8 LIST OF FIGURES Figure Page 1. Model to calculate LTV Examples of question wording for variables vii

9 CHAPTER I. INTRODUCTION One-on-one marketing is not new to American business. In colonial times it was common for a merchant to have direct contact with the customer. This interaction led to trusted bonds between merchants and their customers. The trusted bond between a business and a customer are the foundations of customer relationship management (CRM). CRM is concerned with the creation, development, and enhancement of individualized customer relationships with carefully targeted customers, resulting in maximizing their total customer lifetime value (LTV) (Reinartz, Krafft, & Hoyer, 2004). Companies want to avoid the mistake of not identifying a good customer, and subsequently, not rewarding the customer accordingly. Companies also want to avoid wrongful classification of a low-value customer as a high-value customer and subsequent overspending of resources. The development of a reliable CRM approach aids in the measurement of customer value and therefore reduces the chance of these errors (Reinartz, Krafft, & Hoyer, 2004). The concept of CRM entered the business world in the 1990 s with a promise to change the way businesses interacted with their customers. However, there are some obstacles. CRM is a cumbersome process. It is expensive and difficult to track and 1

10 2 maintain the large database needed to run CRM effectively. However, recent technological advances have greatly improved CRM capabilities. Despite CRM s popularity, there is still confusion about what it is, what it can do, and the best situations in which to use it. When used properly, CRM can allow a company to better understand its valuable customers needs and wants, allowing measurable customer service standards to be created. It identifies the service components important to customers such as an acceptable wait time or time of transaction. The company can then implement customer service standards. Once the standards are in place, analysis can then be conducted to see if, by meeting the standards, customer satisfaction improves. Further research could also explore the relationship between customer satisfaction and customer retention. Database marketing CRM is often confused with database marketing. Although both use databases to guide marketing strategies, the difference is the focus of the marketing. CRM is aimed at determining and influencing the behavior of individuals through one-on-one marketing. Database marketing is aimed at identifying customer segments and markets to them. Customer relationship management evolved in the 1980 s from database marketing and was made popular with mass mailers such as American Express and State Farm Insurance. Both companies used their customer lists to build relationships with their customers after the initial sale, leading to retention and cross sales (Hughes, 2003). Database marketing assumes that through the collection and organization of information about a business, marketing costs can be reduced and profit can increase. Typically, the information is consumer focused: the date of the last purchase, what was purchased, and

11 3 other demographic information. However, an integrated approach would include information about products, suppliers, competitors, and other business areas. As technology became more sophisticated and economical, database marketing became more accessible and practical to businesses. It became possible to store and use information to build lasting relationships. As a result, it became possible to increase sales and profits by promoting cross sales, repeat sales, and upgrades, by computing customer LTV and using it strategically, and by creating customer loyalty programs (Ragusa, 2001). One of the greatest benefits of database marketing is improved customer service. When there is accurate information about the customer, the customer service representative (CSR) is better able to address questions and concerns, since they are provided with the customer s past purchase behaviors (Bean, 1999). Information such as past purchases, times of purchases, amounts of purchases, along with any relevant demographic information about the customer, are available. The unique customer service also allows a special, individualized relationship to develop between the company and the customer. Building a database CRM s success is dependant on an accurate database. The integrity of the data is important. Not taking care to make sure the data are accurate is a major reason why marketing databases fail (Bean, 1999; English, 1998). Business costs of poor data can be significant, and the investment in data quality generates a payback greater than the initial investment. The true challenge of database marketing is the organization and transformation of numerous scattered data into meaningful customer information (Bean, 1999). Building begins with identifying the sources of data, which include transactional

12 4 information, order entry systems, accounting systems, operational manufacturing systems, sales tracking systems, and outside lists. Customer Data Integration (CDI) is an area within data management that can organize various siloed systems into single customer view (McCormick, 2007). One method is through a hub and spoke customer integration model where a central integration point is created into which all source systems will link. Master customer data is stored within the hub such as name, address, date of birth, address, telephone number, etc. A unique customer identifier is given to link the customer to different spokes of data sources. Next, the data must be organized and maintained in a meaningful way. Customer data can change, and it can be difficult to keep the information current and correct. One way this can be done is through the establishment of consistency keys, which make it possible to detect changes in various data sources. This is part of the function of the unique customer identifier. Anytime data from the spokes of the model conflicts with the master customer data in the main hub algorithms are used to create the best match or determine if a new customer record should be created (McCormick, 2007). Consistency key management ensures recognition of the same customer over time (Bean, 1999). Types of databases There are three main types of databases: operational, marketing, and warehouse. Each database is quality controlled by a different department (Hughes, 2003). An operational database is used to process transaction information and general business information, such as sales, shipments, and payments. The IT department often maintains the operational database since it is based on accounting principles. It is balanced to the

13 5 dollar and is externally audited. The operational database contains information only on current customers, and old data is archived. There are no data on prospective customers until they make a purchase. The IT department also oversees the larger database warehouse. The marketing database receives information from the operational database and is managed by the marketing department. It includes information on current, lost, and prospective customers and the company s communication with them. It also contains data from preferences and profiles provided by the customer, a response history from marketing campaigns, and a customer lifetime value. A customer s lifetime value is defined as a measure of the net profitability received from a given customer during their future lifetime as a customer (Hughes, 2003). The warehouse database combines the two databases into one centralized location and is the truly integrated database. CRM evolved from these integrated databases to create an even more individualized relationship for the customer than the non-integrated, multi-database systems that many companies had been using. Theoretical foundation of CRM The key theoretical basis for CRM research is the relationship-marketing literature. It is believed that building and managing ongoing customer relationships delivers the main marketing message (Morgan & Hunt, 1994; Webster, 1992). CRM allows a single view of the customer across all contact channels. Meaning a CSR can pull up a customer in the database and see the entire relationship the customer has with the company. It is important that information coordinates across time and contact channels to manage the entire relationship systematically. When CRM is conceptualized

14 6 at this level, literature suggests four distinct issues must be recognized: (1) building and managing ongoing customer relationships is the essence of the marketing concept (Morgan & Hunt, 1994; Webster, 1992); (2) relationships evolve through distinct phases (Dwyer, Schurr, & Oh, 1987); (3) it is necessary to interact with customers and manage relationships at each stage (Shivastava, Shervani, & Fahey, 1998); and (4) the distribution of relationship value to the company is not homogenous (Mulhern, 1999; Niraj, Gupta, & Narasimhan, 2001). The first assumption in the theoretical approach of relationship management is that managing relationships is beneficial to business (Reichheld & Teal, 1996). For example in the medical industry, a patient s relationship with their doctor is the foundation of the business relationship. The doctor could be an excellent diagnostician but if the patient does not perceive a meaningful relationship with the doctor then they may take their business elsewhere. Therefore managing customer relationships has a direct impact on the business. However, these observations have been qualified by empirical evidence that stresses the importance of moderating effects (Niraj, Gupta, & Narsimhan, 2001; Reinartz & Kumar, 2000). Enablers such as organizational design, appropriate incentives, IT resources, as well as industry, company, or customer structures, may affect the effectiveness of relationship marketing campaigns. For example, in a medical practice it would be difficult to measure the effectiveness of any relationship marketing campaign because the customer-business relationship is heavily weighted by the patient s relationship with their doctor as well as other external factors like the patient s insurance company or the convenience of the office hours. It is

15 7 essential to keep these mitigating factors in mind when evaluating the effectiveness of CRM. The second assumption of CRM is that relationships evolve with distinct phases (Dwyer, Schurr, & Oh, 1987). Relationships cannot be viewed as multiple separate transactions; rather, the interdependency of the transactions creates a dynamic over time (Reinartz, Krafft, & Hoyer, 2004). The first stage of the relationship is the customer acquisition, followed by retention, and finally relationship termination. The customer or company can terminate the relationship at anytime, either intentionally or unintentionally. CRM is a longitudinal process, and the customer relationship must be able to evolve over time. The third assumption in the CRM process is that the recognition of relationship evolution has implications for the company. Companies should interact with customers and manage relationships differently at each stage (Shivastava, Shervani, & Fahey, 1998). A goal of CRM is to manage the different stages of the relationship systematically and proactively. These touch points or key moments of truth are the specific times the company and the customer make contact (Ragusa, 2001). These moments are important and will be addressed in depth later. The final assumption is that the distribution of relationship value to the company is not homogenous (Mulhern, 1999; Niraj, Gupta, & Narasimhan, 2001). An advantage of CRM is that companies are able to measure profitability based on customers, not just product lines, allowing companies to re-examine resource allocations. The most valuable customers frequently do not receive the company s share of attention and resources while the company overspends on marginal customers. CRM proposes that companies define

16 8 different allocations for different tiers of customers, where the customer s value depends on their economic value to the company (Zeithaml, Rust, & Lemon, 2001). CRM process Researchers have given different names to the CRM process, but they all have the same underlying themes. The stages include: Identification of key moments of truth throughout the customer life-cycle. Identification of the ideal value customer. Identification of the gap between what the company currently offers and what the customer values most. Identification of discrepancies among the current and expected services. Identification of core competencies along with enablers required to close the gap. Customer relationship management is valuable in many industries; however, the insurance business is one where it can be most valuable when implemented and used properly. It takes several years before an insurance customer becomes profitable to a company. Therefore, it is more cost effective to focus efforts on customer retention rather than on customer acquisition (Hughes, 2004). The in-depth relationships developed by CRM can help insurance companies to identify and invest in the most valuable customers. This research focuses on the property and casualty insurance industry. Although the following examples will be insurance specific, the general techniques and processes of CRM remain valid across industries. Key moments of truth The key moments of truth are the critical points in which the connection between attitudes and experiences are reinforced or changed (Hughes, 2004). All moments of

17 9 truth must be identified in each phase of the relationship between the customer and the business. These touch points are interactions between the supplier and customer, and many times these are the points where the customer s expectations and preferences may shift under the influence of an event. For example, when a claim is processed and generates minimal disturbance to the customer, the customer s perception of the company may increase. Each essential area of customer satisfaction, such as billing or claims, will reveal a moment of truth (Foss & Stone, 2002). The information can be gained from research or by using brainstorming with groups from all aspects of customer interaction to understand what moments are most critical to the customer.. There are several obvious moments of truth that are important to the customercompany relationship including when the customer receives a bill, when a customer calls the sales line, when a customer goes to the company web site, when a customer calls the company call center with a question or complaint, or when the company contacts the customer in hopes of renewal, upgrade, or cross-sell. Some less obvious moments of truth could include a customer s birthday, a new birth in the customer s family, when a customer moves to a new city, or any time the customer s insurance needs change. The company s performance in moments of truth will determine whether the customer will stay or defect (Foss & Stone, 2002). St. Paul Travelers provides an example of communicating with customers through moments of truth. St. Paul Travelers, based out of St. Paul, Minnesota, supplies commercial and personal property-casualty insurance along with asset management services. Travelers understood the need to develop a touch point program to increase the customers positive moments of truth. The program focused on five annual touch points

18 10 from the agent that varied with the type of insurance the customer had and the length of time the customer had been with Travelers. Agent touch points included: within 60 days of renewal an annual review of the policy would be sent, within the first quarter a thank you card for renewal is sent, in the second quarter a cross-sell postcard is sent, in the third quarter a newsletter is sent, and in the fourth quarter a seasonal greetings card is sent (Hughes, 2004). The Travelers program showed that for each customer, they had to continually determine the appropriate message, the frequency of the messages that the customer wanted, the desired channel, the timing of the message, and the likelihood of defection. Travelers revealed that 65 percent of customers who defected, never talked to an agent before they left, but 80 percent of the customers that talked to an agent during the year did not leave. The importance of having touch points for their customers is revealed in the fact that without the communication with the agent they were losing customers (Hughes, 2004). Customer lifetime value Once the key moments of truth have been identified, it is necessary to determine each customer s lifetime value. Customer lifetime value is a measure of the net profit that the company receives from a given customer during their future lifetime as a customer (Hughes, 2003). Although several LTV models have been developed so far, one generally accepted superior approach does not exist (Jackson, 1992). The following definitions of key customer costs and revenue sources provide a solid background for initial customer lifetime value calculations.

19 11 It is suggested that large and heterogeneous customer groups be separated into homogenous segments that possess different LTV s. In order to create detailed individual LTV s and to ease calculation efforts, each value component should be calculated separately for each customer segment. Then the specific value figures of each group will serve as a basis for the calculation of the individual LTV s. An examination of basic LTV models reveals that the incorporated variables can generally be classified into three categories: retention rate, revenue, and costs (Reinartz & Kumar, 2000). The retention rate refers to the probability that an individual customer will remain loyal to a company, yielding expected revenue and costs within a fixed period of time (Bauer, Hammerschmidt, & Braehler, 2003). The retention rate can be estimated with the help of empirically validated determinants of loyalty, such as customer satisfaction, switching barriers, and the attractiveness of the alternatives. The second category, revenue, can be classified into four sub-categories: autonomous revenue, up-selling revenue, cross-selling revenue, and contribution margins resulting from referral activities of existing customers. These components play a major role in compiling a complete record of the customer s history over the life cycle and are essential to the identification of operative touch points of contact. Autonomous revenue accounts for factors not directly influenced by the company or that are only affected by standard marketing measures like TV advertising. Essentially, it is basic revenue not including targeted measures to increase up-selling and cross-selling. It is usually calculated by means of traditional procedures of demand forecast, e.g., analyses of time sequences. Up-selling revenue is generated by the additional selling of the same product resulting from increased purchase frequency and intensity in long-life relationships

20 12 (quantity effect, i.e., higher purchase amount per transaction and more transactions per period). It also emerges from a price effect, where selling of higher-priced substitutes of the same category to loyal, long-term customers that are less price sensitive (Reinartz & Kumar, 2000). Cross-selling is defined as the selling of complementary products or product categories respectively which might not otherwise have been bought from the company (Reichheld & Sasser, 1990); for example selling homeowners insurance to an automobile insurance customer. The reference value measures margins from new customers acquired through a referral by existing customers. The basic methods for predicting costs, the third category, are those that are commonly used in product-related accounting. The traditional forecast methods have been supplemented by findings about cost reducing effects of long-term customer relationships (Reichheld & Teal, 1996). Acquisition, marketing, recovery, and sales costs must also be included. There are many LTV equations and models, and, as of yet, there is no single calculation that encompasses all the relevant parts of LTV. The following equation (see Figure 1) from Bauer, Hammerschmidt, and Braehler (2003) summarizes many of the essential facets of LTV, including aspects of revenue, costs, and retention rates. Indirectmonetary contributions such as information, cooperation, and innovation value are also included.

21 13 Figure I. Model to calculate LTV. CLV i Lifetime value of customer i (net present lifetime profit) AC i Acquisition costs of customer i r ti Retention rate of customer i in period t AR ti Autonomous revenue of customer i in period t UR ti Up selling revenue of customer i in period t (retention value) CR ti Cross selling revenue of customer i in period t (cross selling value) RV ti Gross contributions from reference activities of customer i in period t (reference value) MC ti Marketing costs for retaining customer i in period t SC Costs for serving the customer i in period t (cost of sales) TC i Termination costs for the relationship with customer i InfoV ti Information value of customer i in period t CoopV ti Cooperation value of customer i in period t InnoV ti Innovation value of customer i in period t d Discount rate appropriate for marketing investments T Length (in years) of the projection period Once the LTV has been established for the customers, it is possible to develop a profile containing characteristics of the most valuable customers. Insurance market research has revealed that ideal customers value clear routes of access, quick responses, prior customer information available at any point of contact, clear documentation and explanations, a feeling of trust, and competitiveness (Foss & Stone, 2002). Valuable customers can give insight into services and standards that they feel are imperative to the insurance experience. Once standards have been established, the gap between what the insurance company currently offers and what the customer wants can be evaluated. If the gap is small, the company is on target with customer expectations. However, if the gap is substantial, the company is not meeting customer expectations and runs the risk of having a customer defect. United Services Automobile Association (USAA) is an insurance company that understands the importance of evaluating the gap between customer service with the current company standards and the customer s expectations. They regularly gauge the discord between current and expected standards. The insurer has maintained an

22 14 extremely low customer defection rate compared to the industry average. Currently USAA has a retention rate of 97 percent (Chordas, 2002). Their high retention rate has been attributed to their superb customer service. When customers defect, USAA surveys them to understand their reasons for leaving. The feedback is then used when reevaluating customer service improvements. Hypothesis development In the insurance industry, knowing a customer s value is especially important. Customer retention is more cost effective than customer acquisition (Hughes, 2004). It takes several years before a customer becomes profitable to an insurance company; therefore, it is imperative customers do not leave the company prematurely. By having an in-depth relationship with its customers through CRM, insurance companies can determine which customers have a high LTV and are worth investment. A targeted and specific marketing approach to its most valuable customers can lead to decreased costs to the company. Taking into account the cost of acquisition and long-term return from the customer, a 10 percent improvement in customer retention can produce a 30 percent increase in pre-tax profitability. In comparison, 10 percent improvement in acquisition only results in a three percent improvement (Benn, 2004). Customers stay with an insurance company when they are satisfied. By meeting service standards, standards the customers themselves set, satisfaction will increase. When there is a discrepancy between the current and expected experiences, it is in the company s best interest to invest resources to eliminate the gap. In this research, I will explore the relationship between meeting customer derived service standards, customer satisfaction, and retention or their stated likelihood to renew.

23 15 Hypothesis 1: Customer identified satisfaction variables under the control of the insurance company will be positively related to the likelihood to renew with the insurer. As customer satisfaction on the identified variables increases, the likelihood to renew will also increase. In addition, insurance companies often reward their most productive agents/agencies with benefits like bonus advertising funds and preferential treatment for their customers. The additional efforts by the insurance company for their high valued agents, keeps them happy and helps them continue to produce quality customers. In this research, I also plan to explore the relationship between customers of high and low value agencies and the differences in agent satisfaction and likelihood to renew. Hypothesis 2: Customers identified as having a Platinum or Gold agent (referred to as customers of high-value agencies ) through the insurance company s FOCUS program will be more influenced by agent satisfaction when choosing to renew than will customers identified as having a non-platinum or Gold agent (referred to as customers of low-value agencies ).

24 CHAPTER II. METHOD Company history In 2004, a regional property and casualty insurer announced the start of an annual customer satisfaction and retention research project. The stated purpose of the research was to achieve the following objectives: explore issues related to performance standards; identify factors that most affect customer satisfaction and retention and the relative importance of each; understand the relative importance of factors influencing selection of an insurance provider; and, profile retention factors and attitudes of personal line customers. The information from this research was to serve as a benchmark for the company s future waves of customer satisfaction and retention. In addition, warehouse information was also included on the agent FOCUS status of each customer interviewed. The FOCUS benefit program was created in The primary purpose of the program is to segment the agency force by performance determined by retention, growth and loss ratio. From which, a rating or focus score is 16

25 17 assigned. A better performance results in a higher rating. Ratings translate to levels: Platinum, Gold, Level 3 and Level 4. The program used to encourage and reward desired agent behavior. Targets adjust every two years to increase the minimum amount of growth per level, decrease the acceptable loss ratio etc. Platinum and Gold agencies are eligible for additional bonus compensation and get more subsidy for reimbursement for marketing and agent training. Internally, services and additional resources are given to Platinum and Gold agencies to help them provide the best service to their customers. Survey procedure The primary objective was to examine and prioritize the current customer service standards and to determine which standards should be retained, which needed to be dropped or modified, and what new standards may be needed to increase customer satisfaction. To achieve this goal, data were collected through in-depth telephone interviews with the company s current personal line (property and casualty insurance products designed for and bought by individuals, including homeowners and automobile policies) customers. The questionnaire was created in conjunction with the insurance company s internal marketing research department and an outside marketing research firm and was defined by past qualitative research and the insurer s predefined needs. The data were collected between October 15 and November 18, The survey was originally timed at 25 minutes; however, demographic questions were dropped to cut the time to 20 minutes. The changes did not interfere with the core standards measures. The final survey contained 18 questions that centered on the company s current customer service standards in the area of billing, claims, and personal lines services. Standards were separated into areas of claims, billing, and personal lines services (e.g.,

26 18 endorsements, new applications, and renewals). An example of survey questions is in figure 2. The full survey can be found in the Appendix. Figure 2. Examples of question wording for variables. Using the same 0 10 scale, with 0 meaning you are completely dissatisfied, and 10 meaning your are completely satisfied, what number would you use to indicate your level of satisfaction with: a. The insurance agent who offers you XX insurance? b. How quickly the agent responds to your calls and questions? c. How quickly XX responds to your calls and questions? d. The ease with which billing is handled? e. The speed at which policy changes are incorporated? Using the 0 10 scale, with 0 meaning you are highly unlikely and 10 meaning you are highly likely, what number would you use to indicate your likelihood to renew your insurance with XX? Participants A total of 506 current personal line customers of the insurer with and without past claims experience were interviewed. Claims experience was defined as customers who had placed a claim after January 1, Respondents were randomly selected from the company s customer database.

27 CHAPTER III. RESULTS In order to test the proposed positive relationship between satisfaction with service standards and the likelihood to renew with the insurer, a multiple regression was conducted. The independent variables consisted of variables identified as important to customer satisfaction through the insurer s previous qualitative research. Only variables that were under the control of the insurer were examined (i.e., variables controlled by the agent were left out). The variables included: Satisfaction with the contacts at the insurance company. Satisfaction with how quickly the insurance company responded to calls and questions. Satisfaction with the ease in which billing was handled. Satisfaction with the speed at which policy changes were incorporated. Satisfaction with how quickly claims were settled. Satisfaction with the ease of the claims process. Satisfaction with the fairness of claim settlements from the insurer. Satisfaction with the advice received from the insurer on ways to reduce problems that might lead to claims. 19

28 20 Satisfaction with the courtesy of people they may have dealt with at the insurance company. Satisfaction with the ease of doing business with the insurance company. Satisfaction with the options you had for how often to pay for your premium. Satisfaction with explanations of premium costs. Confidence that the insurer would take care of you to your satisfaction if you had a claim. The multiple regression method was simultaneous forced entry with all independent variables being entered into to equation model at the same time. Table 1 displays the results of the regression. Table I. Results from likelihood to renew regression of satisfaction variables. Variables Mean Std. r Raw Std. Beta Sig. Tol. Dev. Beta Error Satisfaction contacts at the insurance company Satisfaction with how quickly the insurance company responded to calls and questions Satisfaction with * ease of billing Satisfaction with the speed of policy changes Satisfaction with how quickly claims settled Satisfaction with ease of claims process

29 21 Satisfaction with the fairness of claim settlements from the insurer Satisfaction with the advice received from the insurer to reduce problems that might lead to claims Satisfaction with the courtesy of people they may have dealt with at the insurance company Satisfaction with the ease of doing business with the insurance company Satisfaction with the premium payment options Satisfaction with * explanations of premium costs Confidence that the insurer would take care of you to your satisfaction if you had a claim R 2 Adjusted R 2 F-Value Sig The multiple regression results show a significant overall prediction of the likelihood of the respondents to renew their policy with the insurer, with 53.8 percent of the variance explained by the predictors. All the predictors are correlated with the dependent variable at the.05 level. However, only two variables had significant beta values (satisfaction with the ease in which billing was handled β =.242 and satisfaction with explanations of premium costs β =.145).

30 22 Substantively, the model is shown to be significant. Therefore, satisfaction in the identified service standards can be used to predict a current customer s likelihood to renew with the insurer. When looking at the significant unique contributors influencing their likelihood to renew with the insurer, satisfaction with the ease in which the billing was handled had the greatest influence, and satisfaction with explanations of premium costs the next greatest influence. To address the issues of the multicollinearity in the regression and to reduce the number of variables in the analysis, a factor analysis was conducted. The independent variables from the regression were factor analyzed using principal component analysis with Varimax (orthogonal) rotation. Table 2 displays the results. Table II. Results from factor analysis of independent variables. Factor 1 Factor 2 Factor 3 Communality People service Claims service Premium/policy Satisfaction -- How quickly insurance company responded to calls and questions Satisfaction with contacts at insurance company Satisfaction -- Courtesy of people you may have dealt with at insurance company Satisfaction -- Ease of doing business with insurance company Satisfaction -- How quickly claims were settled Satisfaction -- Fairness of claim settlements from insurance company Satisfaction -- Ease of going through claim process

31 23 Satisfaction -- Advice received from insurance company on ways to reduce problems that might lead to claims Confidence you had that insurance company would take care of you to your satisfaction if you had a claim Satisfaction -- Explanations as to why premium costs where what they were Satisfaction -- Ease with which billing was handled Satisfaction -- Options you had for how often to pay your premium Satisfaction -- Speed at which policy changes were incorporated Eigenvalue % of Total Variance Total Variance % % of Common Variance Kaiser-Meyer-Olkin.909 Measure of Sampling Adequacy. Bartlett's Test of Sphericity Approx. Chi Square df 78 Sig..000 The analysis yielded three independent factors explaining % of the variance for the entire set of variables. Factor 1 was labeled people service due to high loadings by the following items: satisfaction with how insurance company quickly responded to calls and questions; satisfaction with contacts at insurance company; courtesy of people you may have dealt with at insurance company; and ease of doing business with the insurance company. The first factor explained % of the

32 24 variance. Factor 2 was labeled claims service due to high loadings on the following items: satisfaction with how quickly claims were settled; satisfaction with fairness of claim settlements from insurance company; satisfaction with ease of going through claim process; satisfaction that advice received from insurance company on ways to reduce problems that might lead to claims; and satisfaction; and confidence that the insurer would take care of you to your satisfaction if you had a claim. The second factor explained % of the variance. Factor 3 was labeled premium/policy due to high loadings on the following items: satisfaction with explanations to premium costs; satisfaction with ease of billing; satisfaction with premium payment options; and satisfaction with speed of policy changes. The third factor explained % of the variance. Another simultaneous forced entry multiple regression was conducted using the factor scores as the independent variables. Results are shown in table 3. Table III. Results from likelihood to renew regression with factor scores. Variables Mean Std. r Raw Std. Beta Sig. Tol. Dev. Beta Error Factor 1 People *.995 service Factor 2 Claims *.996 service Factor *.999 Premium/policy R 2 Adjusted R 2 F-Value Sig With the factor scores, the multiple regression results show a significant overall prediction of the likelihood of the respondents to renew their policy with the insurer, with 48.0 percent of the variance explained by the predictors. All the predictors are correlated with the dependent variable at the.05 level and have significant beta values.

33 25 Substantively, the model is shown to be significant. Therefore, satisfaction in the identified service standards can be used to predict a current customer s likelihood to renew with the insurer. When looking at the significant unique contributors influencing their likelihood to renew with the insurer, factor 3 (premium/policy) had the greatest influence, followed closely by factor 1 (people service) and factor 1 (claims service). In order to investigate the impact of the FOCUS program model on the likelihood to renew, additional regressions were conducted with respondents divided as customers of high value agents (FOCUS levels Platinum and Gold; n = 114) and customers of low value agents (FOCUS levels 3 and 4; n = 43). Respondents not assigned to a FOCUS agent were removed from the analysis. Different from the previous regression, the independent variables consisted of variables under control of the agent were examined (i.e., variables controlled by the insurer were left out). The variables measured included: Satisfaction with the insurance agent. Satisfaction with how quickly the agent responded to calls and questions. Satisfaction with the ease in which billing was handled. Satisfaction with the speed at which policy changes were incorporated. Satisfaction with the way agent helped with claims Satisfaction with how quickly claims were settled. Satisfaction with the ease of the claims process. Satisfaction with the courtesy of people at agent s place of business. Satisfaction with the ease of doing business with agent. Satisfaction with explanations of premium costs

34 26 The multiple regression method was simultaneous forced entry with all independent variables being entered into to equation model at the same time. Table 4 displays the results of the regression with customers of high value agents. Table IV. Results from likelihood to renew regression of satisfaction variables (customers of high-value agents). Variables Mean Std. r Raw Std. Beta Sig. Tol. Dev. Beta Error Satisfaction with * insurance agent Satisfaction with how quickly agent responded to calls and questions Satisfaction with ease * of billing Satisfaction with the speed policy changes Satisfaction with way agent helped with claims Satisfaction with how quickly claims settled Satisfaction the ease of claims process Satisfaction with the courtesy of people at agent s place of business Satisfaction with ease of doing business with agent Satisfaction with * explanations of premium costs R 2 Adjusted R 2 F-Value Sig The multiple regression results show a significant overall prediction of the likelihood of the customers of high value agents to renew their policy with the insurer, with 33.8 percent of the variance explained by the predictors. All the predictors are correlated with

35 27 the dependent variable at the.05 level. Three variables had significant beta values (satisfaction with the insurance agent β =.209, satisfaction with the ease in which billing was handled β =.207 and satisfaction with explanations of premium costs β =.172). Substantively, this model is shown to be significant. Therefore, satisfaction in the identified agent service standards can be used to predict a current customer s likelihood to renew with the insurer. When looking at the significant unique contributors influencing the high value agents customers likelihood to renew with the insurer, satisfaction with the agent had the greatest influence followed by the ease in which the billing was handled and satisfaction with explanations of premium costs. Table V. Results from likelihood to renew regression of satisfaction variables (customers of low value-agents). Variables Mean Std. r Raw Std. Beta Sig. Tol. Dev. Beta Error Satisfaction with insurance agent Satisfaction with how quickly agent responded to calls and questions Satisfaction with * ease of billing Satisfaction with the speed of policy changes Satisfaction with way agent helped with claims Satisfaction with how quickly claims settled Satisfaction with ease of claims process *

36 28 Satisfaction with the courtesy of people at agent s place of business Satisfaction with ease of doing business with agent Satisfaction with * explanations of premium costs R 2 Adjusted R 2 F-Value Sig The multiple regression results show a significant overall prediction of the likelihood of the customers of low value agents to renew their policy with the insurer, with 47.4 percent of the variance explained by the predictors. All the predictors are correlated with the dependent variable at the.05 level. As in the previous model three variables had significant beta values (satisfaction with the ease of the claims process β =.494, satisfaction with explanations of premium costs β =.263 and satisfaction with the ease in which billing was handled β =.231). Substantively, this model is shown to be significant. When looking at the significant unique contributors influencing the low value agents customers likelihood to renew with the insurer, satisfaction with the claims process had the most influence followed by satisfaction with explanations of premium costs and ease in which billing was handled.

37 CHAPTER IV. DISCUSSION The central aim of this study was to investigate the extent to which variables identified through previous qualitative research would predict likelihood to renew with the insurer. While causality is limited in the models, the company s previous qualitative research supports the inclusion of the variables as predictors. The results from the likelihood to renew regression with insurer controlled variables provided evidence for a positive relationship between satisfaction with the ease in which the billing was handled and satisfaction with explanations of premiums costs. The observed relationship is supported by past studies (Morgan & Hunt, 1994; Webster, 1992). In addition, the results from the analysis on customers of low- and high-value agents support the impact of the FOCUS program on customers likelihood to renew with the insurer. While ease of billing and explanation of premium costs influenced customers of both high- and low-value agencies, satisfaction with the agent was a significant contributor to their likelihood to renew for customers of high-value agencies. These 29

38 30 findings suggest the customer facing benefits from the FOCUS program positively influences the customer s likelihood to renew. It seems that although the customer is unaware of the high-value agency s benefits, the customer has a more positive experience, and in turn, is more satisfied with their agent, leading them to renew. At the time of research, CRM was not implemented at the company due to past failed attempts with database marketing. This study gives adequate support to the usefulness of individualized focus to customers and agents. In the past the company had attempted database marketing, but had difficulty maintaining an accurate customer database. Each area of business and independent agencies had its own database of customer information, but this information was not easily shared with other business units or with the parent insurance company. One customer could be in several databases depending on their policies, and there was not one complete customer database with unique identifiers for each policyholder with all of their demographic and policy information. Customers would call the insurance company with questions and could be transferred several times to different areas before having their question resolved. As a result, customer service satisfaction decreased, and the relationship between the company and the customer was never developed. Since the company was unable to succeed with database marketing, it was unable to explore CRM. Although at the time of this research the company had not implemented CRM, there are definite stages of the process that can be identified. Key moments of truth have been identified through the company s previous qualitative research. All three regressions analyzed showed the importance of touch point opportunities such as the explanation of premiums and billing statements. The satisfaction with these moments of